Absolute address location of variables and functions in Keil C51

1. Variable absolute address positioning

1) When defining a variable, use the _at_ keyword plus the address.

unsigned char idata myvar _at_ 0x40; 

Define the variable myvar at 0x40 of idata. You can find the following line in the M51 file:   

IDATA 0040H 0001H ABSOLUTE ; indicates that a variable is located at the absolute address 0x0040 of idata. 

2) Use KeilC compiler to define variables with absolute addresses, the method is to be checked.

2. Function absolute address positioning

1) Write a function myTest in the program

void myTest(void)

{

　　// Add your code here

} 

2) Use KeilC compiler to locate the function of absolute address, open Project -> Options for Target menu, select BL51 Locate tab, enter in Code:?PR?myTest?MAIN(0x4000) to locate the function myTest to 0x4000 in the program area, and compile again.

3) How to locate multiple functions at once

Similarly, write another function myTest1 in the program

void myTest1(void)

{

　　// Add your code here

} 

　　In the Code: field of the BL51 Locate tab of the Options for Target menu, enter: ?PR?myTest1?MAIN(0x3900), ?PR?myTest?MAIN(0x4000) to locate the function myTest1 at 0x3900 in the program area, define the function myTest at 0x4000 in the program area, and recompile.

The following can be found in the M51 file:

Copy code

3.obj TO Reader RAMSIZE (256) CODE (?PR?MYTEST1?MAIN (0X3900), ?PR?MYTEST?MAIN (0X4000))

　　3665H 029BH *** GAP ***

CODE 3900H 0014H UNIT?PR?MYTEST1?MAIN

　　3914H 06ECH *** GAP ***

CODE 4000H 0014H UNIT?PR?MYTEST?MAIN 

Copy code

4) Function call

The method of directly calling functions in a program will not be explained here. Here we will focus on the method of using function pointers to call functions at absolute addresses.

　　(1) Define the prototype of the function to be called typedef void (*CALL_MYTEST)(void); This is the prototype of a callback function, and the parameter is empty.

　　(2) Define the corresponding function pointer variable CALL_MYTEST myTestCall = NULL;

　　(3) The function pointer variable is assigned to point to the function at the absolute address we located myTestCall = 0x3900; points to the function myTest1

　　(4) Function pointer call

if (myTestCall != NULL)

{

    myTestCall(); // Call the function myTest1 at the function pointer, set the PC pointer to 0x3900

} 

　　Check the bin file generated by the compilation. You can see the content of myTest1 at 0x3900 and the content of myTest at 0x4000.

(5) Other notes: If there is no content in the program space from 0x3000 to 0x3900, when the address pointer of myTestCall points to 0x3800 (between 0x3000 and 0x3900), execution will start from 0x3900. The method of calling a function in AP in Load is similar to this, but the corresponding parameter passing may require another method.

　　Definition of Segment 

　　RSEG is a segment selection instruction. To understand its meaning, you need to understand the meaning of segment.

　　A segment is a storage unit for program code or data objects. Program code is placed in the code segment, and data objects are placed in the data segment. There are two types of segments: absolute segments and relocation segments. Absolute segments are specified in assembly language, and their addresses will not change when connected with L51. They are used to access the i/o of a fixed memory, or to provide the entry address of an interrupt vector. The address of the relocation segment is floating, and its address is determined by L51 when connecting the program module. The segments generated by C51 when compiling the source program are all relocation segments, which have segment names and storage types. Absolute segments do not have segment names.

After saying so much, you may still not understand what a segment means. Don't worry, just keep reading.

For example, you wrote a function void test_fun(void) { ...} in C and saved it in test.c. After compiling it with the compiler, you can see in the .SRC FILE: 

?PR?test_fun?TEST SEGMENT CODE //(Put the function in the code segment) 

When writing this function body:

RSEG ?PR?test_fun?TEST //Select the located code segment as the current segment test_fun:

……//code 

So the expression pattern of the function is: ?PR?function name?file name

The function names are divided into:

1: Function without parameters? PR? Function name? File name

2: Function with parameters?PR?_Function name?File name

3: Re-enter the function?PR?_?function name?file name

For example, you define a global variable

unsigned char data temp1,temp2; 

unsigned char xdata temp3;  

In the test.c file, the compiler will divide each file into 0 or more global data segments, and global variables of the same type are stored in the same segment. So the above will be compiled as follows:

Copy code

RSEG ? DT ? TEST

. temp1: DS 1

. temp2: DS 1

;

RSEG ?XD? TEST

. temp3: DS 1 

Copy code

Copy code

// Below is the representation of the global segment of each type of data

?CO? File name //Constant section

?XD? FILE_NAME //XDATA data segment

?DT? FILE_NAME //DATA data segment

?ID? FILE_NAME //IDATA…..

?BI? FILE_NAME //BIT …..

?BA? FILE_NAME //BDATA….

?PD? FILE_NAME //PDATA…..

Copy code

　　You should understand the meaning of segment after reading this. You may ask, what is the use of this? It is used when you need to write a part of the program in assembly language. Put the function written in assembly language in this file, rename it to xxx.a51, and write it according to the above rules. Just compile it.

　　Now that we know the meaning of segment, let's go back to the usage of SEG. There are two types of segment selection instructions in A51: relocation segment selection instruction and absolute segment selection instruction. They are used to select whether the current segment is a relocation segment or an absolute segment. Using different segment selection instructions will locate the program in different address spaces.

　　1. The selection instruction for relocation segment is: RSEG segment name

It is used to select a previously defined relocation segment as the current segment. The usage is just like the example above, where a function segment is declared first and then the function segment is written.

　　2. Absolute segment selection instruction

Copy code

CSEG [AT absolute address expression] //absolute code segment

DSEG [AT absolute address expression] // internal absolute data segment

XSEG [AT absolute address expression] //External absolute data segment

ISEG [AT absolute address expression] //Internal indirect addressing absolute data segment

BSEG [AT absolute address expression] // absolute bit addressing segment 

Copy code

Here is an example of their usage:

For example, if we write a serial port interrupt program, the starting address is 0x23.

CSEG AT 0X23

LJMP serialISR

RSEG ?PR?serialISR?TEST

. serialISR:  

　　The assembly function uses the variables in the same project C file. For example, if ICFLAG is defined in the C file, the definition in the assembly file is

EXTERN ICFLAG ; define external variables 

　　Define a function, such as

Copy code

CARDATR:

    ...........

    RET

GLOBAL CARDATR 

Copy code

 To call the CARDATR function in the same project file, you should define the function

extern void CARDATR(void); 

C18 specifies the absolute address of data

For example:

#pragma udata overlay RECBUFS =0x190 //200

UINT8 NUMBER;

UINT8 REC_BUF[31];

#pragma udata